davideisinger.com

My personal website
Log | Files | Refs | README

hypercritical-co-uicpgh.txt (9234B)


      1   • [1]Apps
      2   • [2]About
      3   • [3]Archive
      4   • [4]Contact
      5   • [5]RSS
      6 
      7 [6]Hypercritical●
      8 
      9 ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
     10 
     11 I Made This
     12 
     13 January 11, 2024 at 1:51 PM by [7]John Siracusa
     14 
     15 While the utility of [8]Generative AI is very clear at this point, the moral,
     16 ethical, and legal questions surrounding it are decidedly less so. I’m not a
     17 lawyer, and I’m not sure how the many [9]current and future legal battles
     18 related to this topic will shake out. Right now, I’m still trying to understand
     19 the issue well enough to form a coherent opinion of how things should be.
     20 Writing this post is part of my process.
     21 
     22 Generative AI needs to be trained on a vast amount of data that represents the
     23 kinds of things it will be asked to generate. The connection between that
     24 training data and the eventual generated output is a hotly debated topic. An AI
     25 model has no value until it’s trained. After training, how much of the model’s
     26 value is attributable to any given piece of training data? What legal rights,
     27 if any, can the owners of that training data exert on the creator of the model
     28 or its output?
     29 
     30 A human’s creative work is inextricably linked to their life experiences: every
     31 piece of art they’ve ever seen, everything they’ve done, everyone they’ve ever
     32 met. And yet we still say the creative output of humans is worthy of [10]legal
     33 protection (with some fairly narrow restrictions for works that are deemed
     34 insufficiently differentiated from existing works).
     35 
     36 Some say that generative AI is no different. Its output is inextricably linked
     37 to its “life experience” (training data). Everything it creates is influenced
     38 by everything it has ever seen. It’s doing the same thing a human does, so why
     39 shouldn’t its output be treated the same as a human’s output?
     40 
     41 And if it generates output that’s insufficiently differentiated from some
     42 existing work, well, we already have laws to handle that. But if not, then it’s
     43 in the clear. There’s no need for any sort of financial arrangement with the
     44 owners of the training data any more than an artist needs to pay every other
     45 artist whose work she’s seen each time she makes a new painting.
     46 
     47 This argument does not sit well for me, for both practical and ethical reasons.
     48 Practically speaking, generative AI changes the economics and timescales of the
     49 market for creative works in a way that has the potential to disincentivize
     50 non-AI-generated art, both by making creative careers less viable and by
     51 narrowing the scope of creative skill that is valued by the market. Even if
     52 generative AI develops to the point where it is self-sustaining without
     53 (further) human input, the act of creation is an essential part of a life
     54 well-lived. Humans need to create, and we must foster a market that supports
     55 this.
     56 
     57 Ethically, the argument that generative AI is “just doing what humans do” seems
     58 to draw an equivalence between computer programs and humans that doesn’t feel
     59 right to me. It was the pursuit of this feeling that led me to a key question
     60 at the center of this debate.
     61 
     62 Computer programs don’t have rights^[11]1, but people who use computer programs
     63 do. No one is suggesting that generative AI models should somehow have the
     64 rights to the things they create. It’s the humans using these AI models that
     65 are making claims about the output—either that they, the human, should own the
     66 output, or, at the very least, that the owners of the model’s training data
     67 should not have any rights to the output.
     68 
     69 After all, what’s the difference between using generative AI to create a
     70 picture and using Photoshop? They’re both computer programs that help humans
     71 make more, better creative works in less time, right?
     72 
     73 We’ve always had technology that empowers human creativity: pencils,
     74 paintbrushes, rulers, compasses, quills, typewriters, word processors,
     75 bitmapped and vector drawing programs—thousands of years of technological
     76 enhancement of creativity. Is generative AI any different?
     77 
     78 At the heart of this question is the act of creation itself. Ownership and
     79 rights hinge on that act of creation. Who owns a creative work? Not the pencil,
     80 not the typewriter, not Adobe Photoshop. It’s the human who used those tools to
     81 create the work that owns it.
     82 
     83 There can, of course, be legal arrangements to transfer ownership of the work
     84 created by one human to another human (or a legal entity like a corporation).
     85 And in this way, value is exchanged, forming a market for creativity.
     86 
     87 Now then, when someone uses generative AI, who is the creator? Is [12]writing
     88 the prompt for the generative AI the act of creation, thus conferring ownership
     89 of the output to the prompt-writer without any additional legal arrangements?
     90 
     91 Suppose Bob writes an email to Sue, who has no existing business relationship
     92 with Bob, asking her to draw a picture of a polar bear wearing a cowboy hat
     93 while riding a bicycle. If Sue draws this picture, we all agree that Sue is the
     94 creator, and that some arrangement is required to transfer ownership of this
     95 picture to Bob. But if Bob types that same email into a generative AI, has he
     96 now become the creator of the generated image? If not, then who is the creator?
     97 
     98 Where is the act of creation?
     99 
    100 This question is at the emotional, ethical (and possibly legal) heart of the
    101 generative AI debate. I’m reminded of the [13]well-known web comic in which one
    102 person hands something to another and says, “I made this.” The recipient
    103 accepts the item, saying “You made this?” The recipient then holds the item
    104 silently for a moment while the person who gave them the item departs. In the
    105 final frame of the comic, the recipient stands alone holding the item and says,
    106 “I made this.”
    107 
    108 This comic resonates with people for many reasons. To me, the key is the second
    109 frame in which the recipient holds the item alone. It’s in that moment that
    110 possession of the item convinces the person that they own it. After all,
    111 they’re holding it. It’s theirs! And if they own it, and no one else is around,
    112 then they must have created it!
    113 
    114 This leads me back to the same question. Where is the act of creation? The
    115 person in the comic would rather not think about it. But generative AI is
    116 forcing us all to do so.
    117 
    118 I’m not focused on this point for reasons of fairness or tradition. Technology
    119 routinely changes markets. Our job as a society is to ensure that technology
    120 changes things for the better in the long run, while mitigating the inevitable
    121 short-term harm.
    122 
    123 Every new technology has required new laws to ensure that it becomes and
    124 remains a net good for society. It’s rare that we can successfully adapt
    125 existing laws to fully manage a new technology, especially one that has the
    126 power to radically alter the shape of an existing market like generative AI
    127 does.
    128 
    129 In its current state, generative AI breaks the value chain between creators and
    130 consumers. We don’t have to reconnect it in exactly the same way it was
    131 connected before, but we also can’t just leave it dangling. The historical
    132 practice of conferring ownership based on the act of creation still seems
    133 sound, but that means we must be able to unambiguously identify that act. And
    134 if the same act (absent any prior legal arrangements) confers ownership in one
    135 context but not in another, then perhaps it’s not the best candidate.
    136 
    137 I’m not sure what the right answer is, but I think I’m getting closer to the
    138 right question. It’s a question I think we’re all going to encounter a lot more
    139 frequently in the future: Who made this?
    140 
    141 ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
    142 
    143  1. Non-sentient computer programs, that is. If we ever create sentient
    144     computer programs, we’ll have a whole host of other problems to deal with. 
    145     [14]↩
    146 
    147 ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
    148 [15]← Previous
    149 
    150 © 2010-2024 John Siracusa
    151 
    152 
    153 References:
    154 
    155 [1] https://hypercritical.co/apps/
    156 [2] https://hypercritical.co/about/
    157 [3] https://hypercritical.co/archive/
    158 [4] https://hypercritical.co/contact/
    159 [5] https://hypercritical.co/feeds/main
    160 [6] https://hypercritical.co/
    161 [7] https://hypercritical.co/about/
    162 [8] https://en.wikipedia.org/wiki/Generative_artificial_intelligence
    163 [9] https://www.theverge.com/2023/12/27/24016212/new-york-times-openai-microsoft-lawsuit-copyright-infringement
    164 [10] https://en.wikipedia.org/wiki/Copyright
    165 [11] https://hypercritical.co/2024/01/11/i-made-this#fn:1
    166 [12] https://en.wikipedia.org/wiki/Prompt_engineering
    167 [13] https://nedroidcomics.tumblr.com/post/41879001445/the-internet
    168 [14] https://hypercritical.co/2024/01/11/i-made-this#fnref:1
    169 [15] https://hypercritical.co/2023/10/29/apples-blue-ocean